Bypassing Passive PFC In ATX Power Supplies

[Hitto]

I'm planning to remove and bypass the passive PFC transformer/coil on an ATX power supply.

I don't care about reduced efficiency. I'm worried, howewer, about any problem in terms of increased ripple in the primary rail.

Does the removal of the passive PFC has negligible downsides in terms of increased ripple?

Thank you.

[stj]

Re: Bypassing Passive PFC In ATX Power Supplies

i doubt it effects the ripple.

[Hitto]

Re: Bypassing Passive PFC In ATX Power Supplies

But it reduces harmonics, correct? It looks that it may have some benefits in terms of less stress placed to the components placed onto the primary (based onto the following topic: https://www.badcaps.net/forum/showthread.php?t=67663).

[Stefan Payne]

Re: Bypassing Passive PFC In ATX Power Supplies

passive PFC is just a big fat coil.
So its only a bridge you need.

[Hitto]

Re: Bypassing Passive PFC In ATX Power Supplies

That's good to hear.

If you have more informations about any possible drawback (I'm going to upgrade the primary capacitors from 560uF to 820uF and the goal is to combine as many improvements as possible) feel free to write them.

[Dan81]

Re: Bypassing Passive PFC In ATX Power Supplies

You MIGHT experience slight bulging on some PSUs. YMMV depending on how good was the PSU built.

The one I fixed in that thread you mentioned managed to somehow make a Hitachi HP3 cap (820uF 200v if my memory serves me right) slightly bulge. The PSU is still running fine, although I sold the PC I installed the PSU in and bought a GTX750Ti.

[momaka]

Re: Bypassing Passive PFC In ATX Power Supplies

But it reduces harmonics, correct? It looks that it may have some benefits in terms of less stress placed to the components placed onto the primary (based onto the following topic: https://www.badcaps.net/forum/showthread.php?t=67663).

No, that would be APFC which puts more stress on the primary caps.

PPFC (large line-connected inductor) *shouldn't* put more stress on any of the primary components. However, the reasons I say *shouldn't* is because there are cases where PPFC can stress the primary caps. Like all inductors, a PPFC coil may produce significant inductive kickback in the event of sudden power loss. Thus, if you live in an area where you get frequent power outages and power dips (or very dirty power), then it may be possible for the PPFC coil to put more stress on the primary caps. But in general, that is very rare. So I suggest you leave it, as you get added benefits of extra filtering from noise on the line input and better PF.

[Hitto]

Re: Bypassing Passive PFC In ATX Power Supplies

So I suggest you leave it, as you get added benefits of extra filtering from noise on the line input and better PF.

It should also limit inrush current, isn't it?

By the way, this power supply has no NTC. Which should be a practice for increased efficiency but I do not like that.

I am planning to add a 2.5 Ohm NTC: I know that the value is usually higher (8-10 Ohm) but it is just for reducing a bit the inrush current.

The rating of the NTC is 5.5A but I do not know if it refers to the instantaneous current or the operating current.

[budm]

Re: Bypassing Passive PFC In ATX Power Supplies

NTC has nothing to do with efficiency, once the NTC gets hot the resistance will go down but will not be 0 Ohms so you will have power loss on the NTC.
The purpose of BTC is to limit the inrush current when power is first applied to the circuit, once that is over the NTC should be bypassed so it is no longer in the circuit.

[Hitto]

Re: Bypassing Passive PFC In ATX Power Supplies

Passive PFC removed. Just noticed there is a 1uF/250V Mylar between "A" and "B" pins where the PFC was desoldered. And a 22uF/50V capacitor with its negative leg connected to "A". Shall I remove just the Mylar?

[Hitto]

Re: Bypassing Passive PFC In ATX Power Supplies

More informations.

This is a F.S.P. with almost the same design:
https://www.anandtech.com/Gallery/Album/312#3

No mylar but the 22uF capacitor is there.

[Hitto]

Re: Bypassing Passive PFC In ATX Power Supplies

I've found several revision based onto the same design and looked at the differences.

This one is without PFC:

(From: www.hardwaresecrets.com)

The differences:
1) Anti-series diodes
2) 22uF/50V capacitor
3) 1uF/250V Mylar capacitor

Simply bypassing the PFC will at least short the 1uF/250V capacitor, connecting both its legs together as it had, prior to PFC bypassing, one leg connected on the input wire of the PFC and the other one connected on the output wire. A snubber, maybe?

The anti-series diode protects the PFC in case of surges and is also suggested on official PDFs of PFC designs. Dunno about the 22uF/50V capacitors and I would like to know its purpose, if possible, for the sake of learning.


Some informations which will be useful to other members of the forum:
This power supply is quite sturdy and worked good as it was, so I was worried about removing the PFC transformer and do more harm than good.

The PFC was measured and, at least in this case, also acts a huge inductor (rated 26mH with a resistance of 0.20 Ohms) making to my opinion things a bit better for at least the diode bridge and the primary caps.

Not simply a matter of bridging two poles because there are also quite a bunch of components around it. I suggest keeping the PFC and the other components. At least in this design.